Self-driving cars have been tested since 2009, but so far we could not conclude if they are totally safe. A study by National Highway Traffic Safety Administration reveals that 94% car crashes occur due to human error. So, the question arises here, could autonomous cars eliminate accidents caused due to drunk driving, speeding, distraction, and fatigue of human drivers?
To answer this question, an RAND study was revealed yesterday which states that the self-driving cars have to complete “hundreds of millions of miles and sometimes hundreds of billions of miles” to prove their safety rating. Evidently, this would require hundreds of years, while automakers are claiming to put AVs on the roads in a few years time.
If we talk about statistics of 2013, Americans covered around 3 trillion miles in contrast to the Google’s 1.3 million miles which they covered in last 7 years. And mind it these are the most miles any autonomous tech company has covered so far. In an effort to cover 3 trillion miles, 32,028 lives were lost in addition to 2.29 million injuries, while there has been more than a dozen accidents reported for Google autonomous cars, most of them were due to human error. So, to prove that the autonomous cars are safer than humans, we have to drive them for 100 million miles without a casualty. And still then we won’t be able to draw a statistical conclusion.
Seeing that it would be impractical to cover these many miles to approve self-driving car’s safety capabilities, the best way to ascertain their safety is to put them on the road in a controlled setting. Only then we will be able to learn about their real-time advantages and the shortcomings.